6407
are therefore much weaker bases than dialkylhydroxylamines (pK,. = 5.232) or dialkylamino radicals (pK, = 7.0 f 0.533),both of which must be protonated on The absence of any epr signal at H o values between -3.5 and -7.5 is a line-broadening effect which results from a very rapid protonation and deprotonation of the radicals. fast
RzNO*
+ H+ e R,NOH. fast
+
The absence of an epr signal from -3.5 < H o < -7.5 implies that the decrease in a H ~ O Hfor 1H+ with increasing temperature (particularly in 1 : 1 (v/v) CF3COOH-HzS04 in which the proton hfc was not the ketone or alcohol group could have such a profound effect on pK. We therefore attempted to repeat the potentiometric titration on a very pure sample of 2,2,6,6-tetramethylpiperid-4-one N-oxyl. 30 There was no break in the titration curve from pH 8 to 2 and the nitroxide still gave a strong epr signal. It must be concluded that Rozantsev and Gintsberg titrated an impurity. (28) H. Hogeveen, H. R. Gersmann, and A. P. Praat, Recl. Trau. Chim. Pavs-Bas. 86. 1063 (1967). (29) E:G. Rozantsev and E.'G. Gintsberg, Izu. Akad. Nauk SSSR, Ser. Khim., 571 (1965). (30) Purified by slow sublimation at room temperature, mp 40" (lit.31 mp 38"), 100% pure by glc on a 12-ft 10% silicone rubber column at 120". (31) R. Briere, H. Lemaire, and A. Rassat, Bull. SOC. Chim. Fr., 3273 (196.5). (32) T. C.Bissot, R. W. Parry, and D.H. Campbell, J . Amer. Chem. SOC.,79, 796 (1957). (33) R. W. Fessenden and P. Neta, J. Phys. Chem., 76,2857 (1972). (34) The pK. for phenylhydroxylamine protonated on nitrogen is 2.0.35 The estimated pK, for protonation on oxygen is < - 1.74.35 (35) A. Darchen and P. Boudeville, Bull. SOC. Chim. Fr., 3809 (1971).
resolvable at 80") cannot be due merely to the conversion of 1H+to 1. That is, conversion to the unprotonated form, because of a decrease in H o g or shift in the equilibrium constant, would require that the epr signal of 1H+ first broaden and then disappear entirely before that due to 1 developed. However, the epr signal was present at all temperatures and the lines remained sharp as they coalesced. We therefore suggest that the geometry of lH+ at low temperatures differs from that at high temperature^.^^ The decreased interaction between the unpaired electron and the proton at the higher temperatures might, in principle, be due to (i) a more tetrahedral geometry at nitrogen, (ii) a shift of the proton into the nodal plane of' the unpaired electron orbital, or (iii) a rapid inversion of the nitrogen atom (k 2 X lo7 sec-I). The first possibility seems unlikely since aN remains unchanged or decreases slightly as the temperature is raised implying that the geometry of the nitrogen remains unchanged or that it becomes slightly more planar. Even if (ii) or (iii) provide the correct explanation for this phenomenon it is not obvious why lH+behaves differently in concentrated sulfuric acid, nor is it clear why 2H+should behave differently from 1H+in both media. It is to be hoped that future studies of these radicals will answer some of the questions raised by the present work.
-
Acknowledgment. We are grateful to Professor Hoffman for a number of helpful comments and suggestions.
Pteridines. XXVIII. A New, General, and Unequivocal Pterin Synthesis' Edward C. Taylor," Katherine L. Perlman, Ian P. Sword, Margareta S6quin-Frey,2 and Peter A. Jacobi Contributionf r o m the Department of Chemistry, Princeton University, Princeton, New Jersey 08540. Received February 17,1973
Abstract: A versatile new synthetic route to pterins is described. Reaction of an a-ketoaldoximeor a-ketoketoxime with esters of a-aminocyanoacetic acid gives 2-amino-3-alkoxycarbonylpyrazine1-oxides which are cyclized with guanidine to pterin 8-oxides. Deoxygenation of the pyrazine and pterin N-oxides, and the conversion of the latter to 7,8-dihydropterins,are described.
T
he classical synthetic route to pteridines involves condensation of a 4,5-diaminopyrimidine with an a,p-dicarbonyl compound, and it is a fair estimate that more than 90% of all known synthetic pteridines have been prepared by some modification of this condensation reaction. Its usefulness and flexibility (1) (a) Part XXVlI: E. C. Taylor, S. F. Martin, Y. Maki, and G . P. Beardsley, J. Org. Chem., 38,2238 (1973). (b) This work was supported in its initial stages by grants to Princeton University from the National Institutes of Health, Public Health Service (Grants No. CA-02551 and CA-12876),and from the U. S . Army Medical Research and Development Command (Contract No. DA-49-193-MD-2777).This is Contribution No. 1195 in the Army research program on malaria. (2) Postdoctoral Fellow of the Swiss National Science Foundation. (3) For general references on pteridine chemistry, see (a) "Pteridine Chemistry," W. Pfleiderer and E. C. Taylor, Ed., Pergamon Press,
result from the wide diversity of substitution patterns possible with both components and the generally high yield with which the condensation proceeds. The use of a-ketoaldehydes in this so-called Isay pteridine synthesis4 leads, however, to the preferential formation of 7- rather than 6-substituted isomers, while the use of other unsymmetrical a,@-dicarbonylcompounds necessarily gives products of ambiguous structure. 3 ~ 5 Thus, London, 1964; (b) R. C. Elderfield and A. C. Mehta, "Heterocyclic Compounds," Vol. 9, R. C. Elderfield, Ed., Wiley, New York, N. Y., 1967,pp 1-117; (c) W. Pfleiderer, Angew. Chem., Int. Ed. Engl., 3, 114 (1964). (4) 0.Isay, Ber., 39, 250 (1906). ( 5 ) See, for example, (a) R. F. W. Spickett and G. M. Timmis, J . Chem SOC.,2887 (1954); (b) R. Tschesche and G. Sturm, Chem. Ber., 98,851 (1965).
Taylor, et al. 1 Pterin Synthesis
6408
Scheme I
H ~ NHN\NHz + NH~
+
CH,COCHO
----t
0
1
2
(major)
(minor)
in the hay-type synthesis (see Scheme I) of 7-methylpterin (l),the structure of the major product is determined by the preferred condensation of the more reactive of the two carbonyl groups (the aldehyde carbonyl) with the more nucleophilic of the orthosituated amino groups (the 5-amino group), resulting in predominant formation of the 7-methyl isomer 1.3c Depending upon the reaction conditions, this product may be contaminated with the isomeric 6-substituted derivative 2, and removal of this isomeric “impurity” may be extremely difficult. However, the majority of naturally occurring pteridines possess carbon substituents at position 6 rather than at position 7 (consider, for example, the structures of folic acid (3), biopterin (4), neopterin (5), 0
3, R = CH,NH ~ C-O N H C H C H , C H , C O O H
I
COOH
OH
OH
I
I
4, R = C H - CHCH, ( L - e r y t h r o )
OH
OH
I
1
5, R = CH- C H CH,OH ( D - e r y t h r o )
etc.) and the Isay reaction thus produces mainly the “wrong” isomer. Many attempts have been made to devise reaction conditions which would permit reversal of the normal direction of condensation so as to favor formation of the 6-substituted isomer (control of pH, addition of hydrazine, sodium bisulfite, or 2-mercaptoethanol,6a the use of a-hydroxy ketones rather than a,p-dicarbonyl compounds, et^.),^ but it is now generally recognized that a mixture of isomers is probably inevitable and that separation may be either difficult or impossible.6b Attempts to overcome this orientation problem by (6) (a) C. B. Storm, R. Shiman, and S. Kaufman, J . Org. Chem., 36, 3925 (1971). These authors report the formation of pure 6-phenylpterin by the condensation of 2,4,5-triamino-6(1H)-pyrimidinonewith phenylglyoxal at pH 4, but the yield was low (41 %), and the insolubility of the product in this case facilitates isolation of the predominant isomer. (b) An excellent example of this intrinsic deficiency in the hay-type pteridine synthesis is found in the recent work of Archer and Scrimgeour,’ whose various oxidation studies on reduced derivatives of commercial 6-methylpterin were in part vitiated by the finding8 that the latter “compound” was actually a previously unsuspected mixture of the 6 and 7 isomers. (7) M. C. Archer and K. G. Scrimgeour, Can. J. Biochem., 48, 278 (1 ~ 971-1). - . --,.
(8) J. M. Whiteley, J. H. Drais, and F. M. Huennekens, Arch. Biochem. Biophys., 133,436 (1969).
the design of alternate, unequivocal condensation reactions include the reaction of a 4-amino-5-nitrosopyrimidine with an active methylene compound bear~ Boon ing a nitrile group (the Timmis r e a ~ t i o n ) ,the and Leigh synthesis involving condensation of a 4chloro-5-nitropyrimidine with an a-amino aldehyde or ketone, followed by reductive cyclization, lo the condensation of a 4-amino-5-nitrosopyrimidine with phenacyl- or acetonylpyridinium salts, l 1 and the use of pyrazine intermediates, which requires closure of the fused pyrimidine ring as the terminal step in the synthetic sequence. This latter amroach, however, has not enjoyed much use because‘bf difficulties en: countered in preparing suitably substituted pyrazine intermediates. In fact, some of the more versatile types of pyrazines suitable for subsequent conversion to pteridines (i.e., 2-aminopyrazine-3-carboxylic acid derivatives) are actually best prepared by ring cleavage of preformed pteridines. 1 2 c , d , 1 3 Thus, none of these alternate syntheses has proven to be truly satisfactory; a general synthetic route to 6-substituted pteridines which is both unequivocal in the orientation of the 6 substituent, and versatile enough to allow introduction of multifunctional 6 substituents, has not until now been available. We describe in this14 and subsequent papersi6 the development of a new, general, and unequivocal synthesis of pteridines which solves this orientation problem. This new approach possesses the following special features: (a) it readily permits the unambiguous introduction at C-6 of a variety of side chains susceptible to chemical modification, and thus provides a simple entry into the pterin natural products possessing multifunctional C-6 substituents (such as biopterin, sepiapterin, neopterin, folic acid, etc.) ; (b) it permits variation in the nature of the pyrimidine substituents through a choice of the type of cyclization reaction employed in the terminal ring closure ; (c) it produces initially a pteridine &oxide which can function as a useful intermediate for the selective introduction of substituents at position 7, and which upon reduction gives 7,8-dihydropteridines; (d) by suitable modification of the structure of the a-oximino carbonyl compound employed in the initial pyrazine cyclization, it provides as well an unequivocal procedure for the synthesis of the isomeric 7-substituted derivatives; and (e) because every pyrazine intermediate (9) For a bibliography of the Timmis reaction, see T. S. Osdene, ref 3a, pp 65-73. (IO) W. R. BoonandT. Leigh,J. Chem. Soc., 1497(1951). (11) I. J. Pachter and P. E. Nemeth, J . O m .Chem., 28, 1197 (1963). (12) Illustrative examples of this route 6 pteridines may be found in (a) S. Gabriel and A. Sonn, Chem. Ber., 40, 4850 (1907); (b) A. Albert, D. J. Brown, and G. W. H. Cheeseman, J. Chem. Soc., 474(1951); (c) E. C. Taylor, J. A. Carbon, and D. R. Hoff, J . Amer. Chem. Soc., 75, 1904 (1953); (d) E. C. Taylor, R. B. Garland, and C. F. Howell, ibid., 78, 210 (1956); (e) G. P. G. Dick and H. C. S. Wood, J . Chem. Soc., 1379 (1955); (f) E. C. Taylor and W. W. Paudler, Chem. Ind. (London), 1061 (1955); (g) W. B. Wright and J. M. Smith, J . Amer. Chem. Soc., 77, 3927 (1955); (h) E. C. Taylor, 0. Vogl, and P. K. Loeffler, ibid., 81, 2479 (1959); (i) F. Dalacker and G. Steiner, Justus Liebigs Ann. Chem., 660,98 (1962). (13) E. C. Taylor in “The Chemistry and Biology of Pteridines,” a CIBA Symposium, G. E. W. Wolstenholme and M. P. Cameron, Ed., J. and A. Churchill, Ltd., London, 1954, pp 2-34. (14) A preliminary communication on this work has appeared: E. C. Taylor and K. Lenard, J . Amer. Chem. Soc., 90,2424 (1968). (15) (a) E. C. Taylor, K. L. Perlman, Y.-H. Kim, I. P. Sword, and P. A. Jacobi, ibid., 95, 6413 (1973); (b) E. C. Taylor and P. A. Jacobi, ibid., 95, 4455 (1973); (c) E. C. Taylor and R. F. Abdulla, Tetrahedron &err., 2093 (1973); (d) E. C. Taylor and T, Kobayashi, J . Org. Chem., in press.
Journal of the American Chemical Society / 95:I9 j September 19, I973
encountered prior t o the pyrimidine ring annelation step is crystalline, the final pteridines are readily prepared in a high state of purity. In the course of synthetic studies directed toward the preparation of aspergillic acid and derivatives, Sharp and Spring described16 the reaction of alkyl a-aminonitriles (from aldehyde cyanohydrins and ammonia) with a-ketoaldoximes to give 3,5-disubstituted-2-aminopyrazine 1-oxides. It occurred to us that a proper choice of the 3 substituent in such 2aminopyrazine 1-oxides should permit a subsequent ring closure t o a pteridine 8-oxide. Thus, the condensation of a-amino~yanoacetamide‘~with oximinoacetone in glacial acetic acid solution gave 2-amino3-carbamoyl-5-methylpyrazine 1-oxide (6a). The structure of 6a was confirmed by sodium dithionite reduction to 2-amino-3-carbamoyl-5-methylpyrazine (7), identical with an authentic sample prepared by the condensation of methyl glyoxal with aminomalonamidamidine. 2o Oximinoacetophenone reacted analogously with a-aminocyanoacetamide to give 2-amino3-carbamoyl-5-phenylpyrazine 1-oxide (6b). The utility of these readily accessible pyrazine intermediates for pteridine synthesis was demonstrated by condensation with ethyl chloroformate to give the intermediate 2-ethoxycarbonylamino derivatives 8a and 8b, which were cyclized with sodium ethoxide to 6methyllumazine 8-oxide (9a) and 6-phenyltumazine 8-oxide (9b), respectively. Similarly, cyclization of the pyrazines 6a and 6b with triethyl orthoformate in dimethylacetamide gave 6-methyl- and 6-phenyl-4(3H)pteridinone 8-oxide (loa and lob, respectively). These reactions are summarized in Scheme 11. For the purposes of cyclization to pterins (2-amino4(3H)-pteridinones), pyrazine o-aminoesters (Le., 11) would be expected t o be more reactive intermediates than the o-aminocarboxamides 6. We have found that the condensation of ethyl a-aminocyanoacetate with a-ketoaldoximes provides a simple and effective synthesis of these reactive and useful pyrazine intermediates. For example, condensation of ethyl aaminocyanoacetate with oximinoacetone gave 2-amino3-carbethoxy-5-methylpyrazine I-oxide (lla), which was cyclized with guanidine in the presence of sodium methoxide to 6-methylpterin 8-oxide (12a). This latter compound could be reduced with aqueous sodium dithionite in almost quantitative yield t o 6-methyl7,8-dihydropterin (13a), identical with an authentic sample of 13a prepared by the procedure of Boon and Leigh,lo and also by the method of Pfleiderer and Zondler.22 It is of particular interest that 6-methyl7,8-dihydropterin (13a) has been shown to be a substrate for dihydrofolate reductase and has been used as a model for dihydrofolic acid in a number of enzymatic studies. Potassium permanganate oxida-
SchemeI1
(16) W. Sharp and F. S . Spring, J . Chem. SOC.,932 (1951). (17) A. H. Cook, I. Heilbron, and E. Smith, ibid., 1440 (1949). (18) Following completion of this work and publication of our preliminary communication (ref 14), it came to our attention that compound 6a had been prepared independently by the same procedure during studies on the synthesis of new sulfanilamidopyrazines. (19) F. Chillemi and G. Palamidessi, Furmaco, Ed. Sci., 18, 566 (1963). (20) 0.Vogl and E. C. Taylor, J . Amer. Chem. SOC.,81,2472 (1959). (21) (a) A. H. Cook, I. Heilbron, and A. L. Levy, J. Chem. SOC., 1594 (1947); (b) B. Ohta, J . Pharm. SOC.Jup., 68, 226 (1948); Chem. Absrr., 48,4440g (1954). (22) W. Pfleiderer and H. Zondler, Chem. Ber., 99,3008 (1966).
(23) J. M. Whiteley and F. M. Huennekens, Biochemistry, 6, 2620 (1967). (24) We have recently found that benzyl a-aminocyanoacetate (as its methanesulfonic acid salt) also condenses readily with a-ketoaldoximes 1-oxides. to give the corresponding 2-amino-3-carbobenzyloxypyrazine Since these latter pyrazines are insoluble in water, this procedure has the advantage that products are readily isolated by pouring the reaction mixture into water. Reduction and/or cyclization then proceeded normally. (25) W. Pfleiderer and W. Hutzenlaub, Angew. Chem., 77,1136 (1965). (26) D. J. Brown in “Mechanisms of Molecular Migrations,” Vol. 1, B. S . Thyagarajan, Ed., Wiley-Interscience, New York, N. Y . , 1968, pp 209-245.
NC
NOH
II/
H,NCCH
1
0
NHZO
\\
+
1
C-CH,
I
+
CH
N
Y=cH;)’ I 0-
0
I
0-
8
6
1 0
0
0-
0-
10
9
a,R=CH3 b,R = CSH,
tion of 13a proceeded quantitatively to give 6-methylpterin (2 14a), identical with an authentic sample.’O In analogous fashion, a number of 2-amino-3-carbethoxypyrazine I-oxides (llb-d) were prepared by condensation of ethyl a-aminocyanoacetate with the appropriate a-ketoaldoximes (see Scheme 111).2 4 The possibility that the pterin 8-oxides 9, 10, and 12 might exist as the tautomeric 8-hydroxy derivatives (Le,, 15) appears to be excluded by comparison of the nmr spectra of the N-oxides with the spectra of the deoxygenated pteridines (for details see Experimental Section). It should be noted, however, that Pfleiderer and Hutzenlaub, in connection with their preparation of 3-methyl-6-phenyllumazine 8-oxide, considered both tautomers as apparently probable, and we therefore record some additional experiments with 12a which convincingly exclude the 8-hydroxy tautomer 15 as a major contributor (see Scheme IV). Thus, 6-methylpterin 8-oxide (12a) was methylated with dimethyl sulfate at pH 7-8 t o give a monomethyl derivative which was shown conclusively t o be the 3methyl derivative 16 by subjecting it to a base-catalyzed Dimroth rearrangement26 to give 2-methylamino-6-
Taylor, et al.
Pterin Synthesis
6410 Scheme 111 /
0
NH2
C,H,OOCCH I
\\
+
I C-R
NC
-
0
II
C 2 H , 0 C x N rR H2N
Y+
N/cH I
0-
OH
11 1,guanidine
0
0
tures are correctly represented as 8-oxides and not as 8-hydroxy derivatives ( i e . , 15). Many of the a-ketoaldoximes employed in the above reactions were prepared by simple nitrosation of the corresponding methyl ketones (i.e., acetone, acetophenone). It should be noted, however, that direct oximation is ambiguous when the carbonyl group is flanked by unsymmetrical active methylene groupings. For example, nitrosation of mesityl oxide would be expected and has been claimed2’ t o give the a-ketoaldoxime 18 (see Scheme V). We have now Scheme V 0
II
0
(CH,),C==CHCCH,
++
18
Jt
13
12
0
II
CH&CHzCCH,
__*
II II
CH&- C-CCH,
It
0
CH? 19
14 (2, R = CH)
1
R = CH,
b, R = CH,CH,CH, E, R = CH=CHC,H5 d, R = C,H,
0
0 guanidine
Scheme IV 0
H ~HNN-JIpEp~H , CCH,
7’
0
0-
[H2
21
016
0
0
0-
OH 15
17
methyl-4(3H)-pteridinone 8-oxide (17). This Dimroth rearrangement was conveniently followed by nmr spectroscopy; 16 exhibits two methyl signals at 2.36 (CB-CHa) and 3.30 ppm (Na-CH3). As the Dimroth rearrangement proceeds, the latter (N-methyl signal) is seen to move upfield to a terminal position of 2.83 ppm (the 2-methylamino grouping), while the 2.36 ppm signal remains unchanged. The percentage of Dimroth rearrangement at any given time is readily determined by integration. The uv spectra of 12a and of 16, both as neutral molecules and as cations, are essentially identical, as are their low pK, values (the higher pK, value for 16 could not be determined because of the intervening Dimroth rearrangement). Since the position of the C-7 proton signal in all of the pteridine 8-oxides examined remains essentially unchanged, and is the same as that exhibited by the C-7 proton signal in 12a, we conclude that their strucJournal of the American Chemical Society
0
HON HONO
II
a,
II
(CHJ2C=CHCCH=NOH
95:19
II
CIH,OC
X)2H3 It
H*N
It
0-
CH2
20
found, however, that this structural assignment is in error, and that the nitrosation product of mesityl oxide is in fact 2-methyl-3-oximinopent- 1-en-Cone (“isonitromesityl oxide”) (19). z8 The structure of 19 was confirmed not only by its definitive nmr spectrum (see Experimental Section) but by its cyclization with ethyl a-aminocyanoacetate t o 2-amino-3-carbethoxy5-methyl-6-isopropenylpyrazine 1-oxide (20). The structure of this latter compound was unambiguously confirmed by examination of its nmr spectrum (see Experimental Section). Condensation of 20 with guanidine in the presence of sodium methoxide then gave 6-methyl-74sopropenylpterin 8-oxide (21). Since analogous difficulties may be anticipated on nitrosation of other ketones,29 we prefer the general synthetic route to a-ketoaldoximes outlined below in Scheme VI, which proceeds in high yield and is unambiguous. The p-keto esters which were not commercially available were prepared either by condensation of the appropriate methyl ketone with diethyl carbonate3O or by reaction of alkyl Grignard reagents with ethyl cyanoacetate. 3 1 Extensions of this new, general, and unequivocal approach to pteridines leading to syntheses, inter alia, (27) L. Claisen and 0. Manasse, Ber., 22,526 (1889). (28) Mesityl oxide is known to be in equilibrium with its proto-
tropic isomer “isomesityl oxide:” F. H.Stross, J. M. Morgen, and H. deV. Finch,J. Amer. Chem. Soc., 69,1627(1947). (29) See, for example, R. W. L. Clarke, A. Lapworth, and E. Wechsler, J . Chem. Soc., 93, 30(1908). (30) V. H. Wallingford, A. H. Homeyer, and D. M. Jones, J . h e r . Chem. Soc., 63,2252 (1941). (31) G . W. Anderson, I. F. Halverstadt, W. H.Miller, and R. 0. Roblin, Jr., ibid., 67,2197(1945).
September 19, 1973
6411 for 3 hr. The yellow sodium salt of the lumazine was collected by Scheme VI filtration and dissolved in water, and hydrochloric acid was added 0 0 0 to p H 3-4. Filtration gave 0.15 g ( 8 6 7 3 of white crystals of 6II II H,O 11 methyllumazine %oxide: mp >320°; nmr ( C F K O O H ) 6 2.25 RCCH,COC,H, RCCH,COOH (3, s, C6-CHa), 8.46 (1, s, Ci-H); uv A:", nm (log E ) 248 sh (4.40), 268 (4.47), 290 sh (4.08), 385 (3.99). 1HOYO 6-Phenyllumazine 8-Oxide (9b). Reaction of 2-amino-3-carbamoyl-5-phenylpyrazine 1-oxide with ethyl chloroformate in 0 n xylene solution, as described above, gave 2-ethoxycarbonylamino3-carbamoyl-5-phenylpyrazine 1-oxide (8b) (84 yield), which was A 11 cyclized to 6-phenyllumazine 8-oxide by heating with sodium RCCH=NOH RCCCOOH methoxide in methanol: yield, 90%, mp >320°; nmr (CFKOOH) II 6 7.30 (3, m), 8.00 (2, m, C6-C6Hs), 8.93 (1, s, C r H ) ; uv Ai: nm NOH (log E ) 248 (3.88), 295 (4.14): 400 (3.36). of pteroic acid, 6-hydroxymethylpterin, folic acid, 6-Methyl-4(3H)-pteridinone8-Oxide (loa). A solution of 0.60 g 1-oxide in 5 ml of triof 2-amino-3-carbamoyl-5-methylpyrazine aminopterin, methotrexate, and the pteridine natural ethyl orthoformate and 10 ml of D M F was heated at 140" (oil products possessing multifunctional C-6 substituents bath) with stirring for 16 hr. The precipitated solid was collected (z',e., biopterin) are described in further papers in this by filtration, washed well with water, ethanol, and ether, and dried series. to give 0.41 g (64%) of 6-methyl-4(3H)-pteridinone 8-oxide: mp >320"; nmr (CFKOOH) 6 2.48 (3, s, C6-C&), 8.66(1, s), 8.30 Experimental (1, s, C?-H and CrH); uv Ai: nm (log E ) 228 (4.05), 225 sh (4.05): 265 sh (4.13), 272 (4.13); 300 sh (3.86), 345 (3.86), 358 2-Amino-3-carbamoyl-5-methylpyrazine1-Oxide (6a). T o a so(3.86). lution of 3.00 g of a-aminocyanoacetamidel7 in 10 ml of glacial 6-Phenyl-4(3H)-pteridinone8-Oxide (lob). A solution of 3.30 g acetic acid was added 3.00 g of oximinoacetone in 10 ml of glacial 1-oxide in a mixture of of 2-amino-3-carbamoyl-5-phenylpyrazine acetic acid. An exothermic reaction ensued and some external 5 ml of triethyl orthoformate and 10 ml of dimethylacetamide was cooling was necessary during the first 10 min to maintain the heated under reflux with stirring for 12 hr. The resulting sustemperature at 20-25". Yellow crystals started to separate after pension of white crystals was cooled and filtered, and the collected 1G15 min. The mixture was stirred at room temperature for 16 solid was washed well with water followed by ethanol and dried to hr, and the yellow crystals were collected by filtration, washed with give 0.25 g (80%) of cream-colored crystals: mp >320"; nmr cold acetic acid followed by ether, and dried to give 3.00 g ( 6 2 z ) of (CF3COOH) 6 7.66 (3, m), 7.16 (2, m, C6-c&), 8.45 (1, s, C?-H), 2-amino-3-carbamoyl-5-methylpyrazine1-oxide, mp 218-219" nm (log e ) 225 (4.23), 295 (4.44), 320 9.10 (1, s, Ci-H); uv ::A: (lit.I9 mp 235-236" from water). Recrystallization from ethanol sh (4.16), 370 (3.89). or ethyl acetate gave bright yellow needles, but without change in 2-Amino-3-carbethoxy-5-methylpyrazine1-Oxide (lla). A soluthe melting point. The compound gave a deep blue ferric chloride tion of 4.35 g of oximinoacetone and 15.00 g of ethyl a-aminocyanotest in aqueous solution: nmr (CFICOOH) 6 2.1 (3, s, Cs-CH,), acetate tosylate in 20 ml of absolute methanol was stirred for 24 hr nni (log e ) 228 (4.05), 246 (4.24), 378 7.83 (1, s, C6-H); uv at 35" (care should be exercised that the temperature does not rise (3.88). higher than 35"). The clear brown reaction solution was then 2-Amino-3-carbamoyl-5-methylpyrazine (7). T o a solution of evaporated under reduced pressure and the residual brown oil 1-oxide in 7 ml of 0.30 g of 2-amino-3-carbamoyl-5-methylpyrazine triturated with 100 ml of water and then extracted with 3 X 300 ml boiling water was added portionwise 3.50 g of sodium dithionite. of ethyl acetate, followed by 1 X 300 ml of chloroform. The The mixture was refluxed for 2 hr, the resulting suspension was combined extracts were dried over anhydrous sodium sulfate, and cooled, and the precipitate was collected by filtration, washed with the solvents were evaporated under reduced pressure to give 4.85 g water, and dried to give 0.21 g ( 7 8 z ) of pale yellow crystals of (4973 of pale yellow needles, mp 132-133.5". An additional 2-amino-3-carbamoyl-5-methylpyrazine, mp 203-204". The mate0.80 g (total yield 5.65 g, 58%) of product was obtained by conrial gave a negative FeC13 test and was identical in all respects with centration of the above aqueous layer to 100 ml, followed by conan authentic sample.20 tinuous extraction with methylene chloride, drying of the extracts 2-Amino-3-carbamoyl-5-phenylpyrazine1-Oxide (6b). To a sus(Na2S04),evaporation, trituration of the brown oil with isopropyl pension of 1.50 g of oximinoacetophenone in 10 ml of glacial acetic ether, and recrystallization of the resulting yellow crystals from acid was added a solution of 1.00 g of a-aminocyanoacetamide in ether: nmr (CDCl,) 6 1.40 (3, t), 4.5 (2, q , CZH& 2.43 (3, s, C55 ml of glacial acetic acid, and the resulting clear solution was CHI), 8.13 (1, s, C&), 7.3 (2, br, G-NH?); uv: : A nm (log e ) 226 stirred for 12 hr at room temperature. The resulting slurry of (4.05), 246 (4.29). 378 (3.94). crystals was diluted with 50 ml of water and filtered, and the col2-Amino-3-carbethoxy-5-rl-propylpyrazine 1-Oxide (llb). This lected solid was washed well with water followed by ethanol and dried to give 0.75 g (32 %) of 2-amino-3-carbamoyl-5-phenyl- compound was prepared in 52% yield from 15.00 g of ethyl aaminocyanoacetate tosylate and 5.75 g of oximinomethyl-npyrazine 1-oxide, mp 28G282". Recrystallization from D M F gave propyl ketone, as described above for the preparation of l l a : long yellow needles with the same melting point: nmr (DMSO) yellow, silky needles, mp (from ether) 89-90'; nmr (CDCls) nm 6 7.50 (3, m). 8.23 (2, m, cj-c&j), 9.16 (1. S, Cs-H); UV 6 1.41 (3, t), 4.41 (2, q, C1H5),0.95 (3, t), 1.70 (2, m), 2.66 (2, t, (log E ) 245 sh (4.08), 280 (4.38), 390 (3.81). nm (log E ) 228 (4.08), 252 C3Hi), 7.96 (1, s, C6-H); uv 6-Methyllumazine 8-Oxide (9a). A mixture of 0.34 g of 2-amino(4.30), 380 (3.91). 3-carbamoyl-5-methylpyrazine1-oxide and 0.54 g of ethyl chloro2-Amino-3-carbethoxy-5-styrylpyrazine 1-Oxide (llc). A soluformate was heated under reflux with stirring for 4 hr. Progress tion of 18.00 g of ethyl a-aminocyanoacetate tosylate and 10.50 g of the reaction could conveniently be followed by testing aliquots of styryl oximinomethyl ketone in 60 ml of 2-propanol was stirred of the reaction mixture with ferric chloride; the starting material at room temperature for 16 hr. The yellow crystalline solid which gives a strong positive (deep blue) test, while the 2-ethoxycarbonylseparated from the dark solution was collected by filtration, washed amino derivative ( d e itifru) gives a negative test. The reaction well with 2-propanol, and dried to give 4.90 g (29 %) of l l c , mp 155mixture was then cooled and the white crystals were filtered off and 156" The analytical sample was prepared by recrystallization from dried to give 0.35 g (85%) of crude 2-ethoxycarbonylamino-3ethanol without change in the melting point: nmr (CDC13) 6 carbamoyl-5methylpyrazine 1-oxide (sa). 1.45 (3, t), 4.50 (2, q, CzHj), 8.43 (1, S , Ctj-H), 7.35 ( 5 , m, CtH;), T o a solution of 0.03 g of sodium dissolved in 10 ml of methanol was added 0.19 g of crude 2-ethoxycarbonylam~no-3-carbamoyl-5- 6.96 (1, d), 8.26 ( 1 , d, J = 17 Hz, CH=CH trans); uv A%:6oH nm (log E ) 315 (4.49), 402 (3.75). methylpyrazine 1-oxide, and the mixture was heated under reflux 2-Amino-3-carbethoxy-5-phenylpyrazine1-Oxide (lld). A mixture of 15.00 g of ethyl a-aminocyanoacetate tosylate and 7.45 g of (32) Melting points are uncorrected and were determined on a oximinoacetophenone in 30 ml of absolute methanol was stirred Thomas-Hoover apparatus. Infrared data refer to Nujol mull spectra at 35" for 24 hr. A clear solution resulted after 5 min, a yellow taken on a Perkin-Elmer 237B grating infrared spectrometer. Nmr crystalline solid started to separate after about 2 hr, and the reaction data were obtained on a Varian A-60A instrument, using TMS as mixture practically solidified after 2.5 hr. Filtration gave 7.48 g internal standard. Elemental analyses for all new compounds were in (60%) of yellow crystals, mp 135-137'. The filtrate was concenagreement with the assigned structures to within 0.3 Z. Results of these analyses were made available to the editor. trated under reduced pressure, and the residual brown oil was dis-
-
-
II
+ -
Azo"
Taylor, et al.
Pterin Synthesis
6412 ethanol, and dried; yield, 0.08 g (90%), mp >320°. The uv, nmr, solved in 50 ml of ethyl acetate and transferred to a separatory and ir spectra of this compound were identical with those of an funnel, and 100 ml of water and 250 ml of ethyl acetate were added. Its measured authentic sample of 6-methyl-7,8-dihydr0pterin.~~,~2 The aqueous layer was separated and extracted again with 2 X 300 pK, values [4.17 (~k0.03)and 10.85 (+0.03)] were identical with ml of ethyl acetate and 1 X 300 ml of chloroform, the combined those reported by PAeiderer and Zondler22 but differed slightly from organic extracts were evaporated to dryness, and the residue was crystallized from 2-propanol or ethanol. This procedure gave an those reported by Whiteley and Huennekens (3.2 and 10.6).23 The reduction could also be effected (42% yield) with hydrogen and additional 1.78 g of product: total yield 9.26 g (74%); nmr (CDRaney nickel catalyst (room temperature, 50 psi of hydrogen). Clr) 6 1.5 (3, t), 4.5 (2, q, G H d , 7.9 (2, m),7.43 (3, m,Ca-CeHd, 6-Methylpterin (2 14a). This compound was readily prepared 8.86 (1, s, c6-H); uv n m (log e) 245 sh (4.09), 278 (4.41), 392 in 9 4 z yield by stirring a solution of 6-methyl-7,8-dihydropterin (3.23). in 0.1 M potassium permanganate solution (made alkaline by addi2-Amino-3-carbethoxy-5-methyl-6-isopropenylpyrazine1-Oxide tion of a small amount of 0.1 N sodium hydroxide) at room tem(20). This compound was prepared in 51% yield from ethyl a-aminocyanoacetate and 2-methyl-3-oximinopent-l-en-4-one perature for 10 min, bubbling in sulfur dioxide to dissolve the precipitated manganese dioxide. and filtering off the precipitated (19),*7 in acetic acid as solvent: mp (from ether) 121-122'; nmr bright yellow, crystalline 6-methylpterin, identical with an authentic (CDClJ 6 7.35 (2, br, G-NHz), 1.43 (3, t), 4.53 (2, q, CZHS),2.10 sample. lo (3, t, J = -2 Hz, CHsC=), 2.50 (3, s, Cb-CHs), 5.60 (1, m), 6-Pheny1-7,S-dihydropterin (13d). A solution of 0.15 g of 6nm (log e) 233 (4.17), 253 (4.25), 287 5.20 (1, m, =CH*); uv phenylpterin 8-oxide in 7 ml of 0.5 N sodium hydroxide containing sh ( 3 3 9 , 3 7 8 (3.94). 0.55 g of sodium dithionite was heated under reflux for 15 min, 6-Methylpterin 8-Oxide (12a). To a methanolic solution of cooled, and acidified to pH 1-2 with hydrochloric acid. The guanidine (prepared by adding 5.60 g of guanidine hydrochloride yellow solid which precipitated was collected by filtration and to 240 ml of methanol containing 2.80 g of sodium, and removal of redissolved in 6 ml of hot 0.5 N sodium hydroxide, and the solution the precipitated sodium chloride) was added 8.00 g of 2-amino-3again acidified to give 0.14 g (88 %) of 6-phenyl-7,8-dihydropterin carbethoxy-5-methylpyrazine 1-oxide. An orange precipitate hydrochloride, mp >320". gradually formed. The mixture was stirred at room temperature Oxidation of this material with potassium permanganate, as for 30 min, the methanol removed by evaporation under reduced described above, gave 6-phenylpterin (14d), identical with an pressure, and 150 ml of D M F added to the residue. The resulting authentic sample.33 solution was heated under reflux for 4 hr and cooled, and the 3,6-Dimethylpterin 8-Oxide (16). Dimethyl sulfate (4 ml) was yellow precipitate was collected by filtration and washed with added slowly and with stirring over a period of 20 min to a soluethanol. The crude product was dissolved in hot water and hytion of 1.00 g of 6-methylpterin 8-oxide in 80 ml of 0.1 N sodium drochloric acid added to pH 3-4. Filtration then gave 5.50 g hydroxide. The pH of the mixture was maintained at 7-8 by (70%) of pure 6-methylpterin 8-oxide: mp >320"; nmr (CFImax slow addition of 1 N sodium hydroxide (approximately 30 ml were COOH) 6 2.36 (3, S, Ca-CHs), 8.50 (1, S , CT-H); uv XDN"82 necessary). The initially clear solution began to deposit yellow nm (log E) 270 (4.39, 290 sh (4.10), 378 (3.85); Xkt,'3 nm (log E) crystals after about 30 min. After 3 hr of stirring, the mixture 260 (4.54), 287 sh (3.92), 387 (3.94); nm (log c) 257 was chilled to 0" and filtered, and the collected yellow solid was (4.29), 287 sh (3.93), 352(3.73). washed with ether, then with cold water, and dried: yield, 0.68 g The following pterin 8-oxides were prepared in the same manner (63%), mp >300"; nmr (CF3COOH) 6 2.36 (3, s, C6-CH3), 3.30 from guanidine and the corresponding 2-amino-3-carbethoxy-5(3, s, N3-CH3), 8.40 (1, s, C7-H); uv- XkHp.33nm (log e) 270 substituted-pyrazine 1-oxides. nm (log e) 257 (4.19), (4.25), 292 sh (3.98), 382 (3.81); A::;-*' 6-n-Propylpterin 8-Oxide (12b): 70 % yield, mp >320"; nmr 287 sh (3.99), 352 (3.62). (CF3COOH) 6 0.58 (3, t), 1.30 (2, m), 2.50 (2, m, C3H7), 8.16 (1, 2-Methylamino-6-methyl-4(3H)-pteridinone 8-Oxide (17). A s, C7-H); uv ::)A: nm (log E) 217 (4.01), 248 (3.90), 276 suspension of 0.20 g of 3,6-dimethylpterin 8-oxide in 4 ml of 2 N (4.31), 300 sh (3.99), 382 (3.70). sodium hydroxide was boiled for 5 min. The pterin quickly dis6-Styrylpterin 8-Oxide (12c): 71% yield, m p >320°; nmr solved to give a clear orange solution, from which the yellow sodium ( C F K O O H ) 6 6.7 (1, d), 7.3 (1, d, CH=CH); 7.1 (5, m, C a s ) , salt of the Dimroth rearrangement product rapidly separated. nm (log c) 245 (4.03), 280 sh 7.70 (1, S , CI-H); UV The suspension was cooled and filtered, and the collected sodium (4.27), 320 (4.53), 415 ( 4 . 0 4 y salt was dissolved in 5 ml of hot water. Acidification with 4 drops 6-Phenylpterin 8-Oxide (12d): 65% yield, mp >320°; nmr of concentrated hydrochloric acid (to pH 2-3) resulted in the (CF3COOH) 6 7.33 (3, m), 7.66 (2, m, CeHr), 9.00 (1, s, C7-HI; uv ~ 0 . 1 . Vh a O H nm (log E) 283 (4.73), 320 sh (4.29), 397 (4.15). precipitation of yellow crystals which were collected by filtration, mar washed with cold water followed by ethanol, and dried to give 6-Methyl-74sopropenylpterin 8-Oxide (21): 86 % yield, mp 0.18 g (90%) of 2-methylamino-6-methyl-4(3H)-pteridinone8>320°; nmr (CF3COOH) 6 1.70 (3, t, J = -2 Hz, CHaC=), 2.33 oxide: mp >300°; nmr (CFICOOH) 6 2.36 (3, s, CrCHs), 2.83 o H (log (3, s, C6-CH3), 5.00 (1, d), 5.46 (1, d, CH2=); uv X ~ ~ ~ b "nm (3, s, CH3NH), 8.40 (1, s, C7-H); uv X~,:~"""" nm (log e) 265 E) 262 (4.53), 290 sh (3.93), 385 (4.02). (4.51), 295 sh (3.90), 400 (3.90). 6-Methyl-7,S-dihydropterin (13a). T o a suspension of 0.10 g of 6-methylpterin 8-oxide in 10 ml of boiling water was added 1.00 g of sodium dithionite. The resulting solution was stirred at room (33) Prepared by the condensation of 2,6-diamino-5-(p-nitrophenylazo)-4(3H)-pyrimidinone with the morpholine enamine of phenyltemperature overnight and the colorless solid which had separated acetaldehyde; this work will be reported independently. was collected by filtration, washed well with water followed by
Journal of the American Chemical Society
95:19
September 19, 1973
